The present invention relates to a switch network for multichannel installations, such as those found in telecommunications satellites, between the antenna ports and the amplifiers, for example.
Such networks must connect one of N input ports (antenna ports, in the example considered later) and one of t output ports (amplifier ports, in the example considered later).
To be more precise, the network must provide P such connections, with P less than or equal to both N and T. This is because there may be redundancy, in the construction of the satellite, both with regard to the antennas and with regard to the amplifiers. In determining each of the P connections between one of the elements of a set consisting of P out of N input ports and one of the elements of a set consisting of P out of T output ports, it is sufficient for each of the P input ports to be connected to a different one of the P output ports. If one of P output ports has to be replaced by another of the T-P previously unused output ports, the switch network must be such and controlled in such a way that only a limited number of connections already set up are modified.
Numerous switch networks meeting these demands are known per se, and those disclosed in patent document EP-0599734-A1 will be cited by way of example, comprising individual 4-pole switches having two opposite ports which can be connected to one another, or each to an adjacent port to the right, or each to an adjacent port to the left. Such networks include a relatively large number of such switches. Because of the technology employed, the switches are heavy and bulky.
The invention starts from switches using a different technology, illustrated by patent documents EP-0670579-A1 and EP-0211541-A2. The switch from the first of these two documents comprises two terminals between which is a movable contact piece driven by an electromagnet. The switch from the second document comprises a central terminal surrounded by a circle of radial terminals, a contact piece being provided between the central terminal and each of the radial terminals. Networks such as the one from document EP-0599734-A1 mentioned above could use such switches. The advantages and restrictions resulting from the change of technology would thereby be obtained, but the drawbacks resulting directly from the relatively high number of switches in a network characterized by given numbers N/PIT would remain.
The invention is aimed at a solution which substantially reduces the remaining drawbacks just mentioned, that is to say a switch network for connecting one of N input ports and one of T output ports, said network comprising an arrangement of switches and being characterized, according to the invention, in that each switch comprises at least one first terminal to which a first port of a first type (an input port, for example) is connected at least one second terminal, to which a second port of a second type (an output port, in this example) is connected, and a contact bridge provided with an actuation device and which on command makes a connection between said first and second terminals, and in that said network comprises a plurality o f net work terminals each constituting at least one of said first terminals for a first switch of said network and one of said first terminals for a second switch of said network.
Said plurality of network terminals preferably comprises a matrix with one of said terminals at each node.
In one embodiment, said plurality of network terminals comprises at least two matrices with one of said terminals at each node, network terminals of one matrix being connected to network terminals of another matrix.